A Note on Slepian-Wolf Bounds for Several Node Grouping Configurations

TL;DR

This paper explores how grouping nodes by correlation and simplifying structures with Markov models affect Slepian-Wolf bounds, making distributed source coding more practical for multi-node wireless sensor networks.

Contribution

It introduces two practical scenarios for node grouping and correlation modeling, simplifying the Slepian-Wolf bounds for multi-node sensor networks.

Findings

Grouped nodes have more restrictive bounds

Simplified models make bounds easier to compute

Practical scenarios improve applicability of Slepian-Wolf coding

Abstract

The Slepian-Wolf bound on the admissible coding rate forms the most fundamental aspect of distributed source coding. As such, it is necessary to provide a framework with which to model more practical scenarios with respect to the arrangement of nodes in order to make Slepian-Wolf coding more suitable for multi-node Wireless Sensor Networks. This paper provides two practical scenarios in order to achieve this aim. The first is by grouping the nodes based on correlation while the second involves simplifying the structure using Markov correlation. It is found that although the bounds of these scenarios are more restrictive than the original Slepian-Wolf bound, the overall model and bound are simplified.